Unsupervised Process Monitoring and Fault Diagnosis with Machine Learning Methods

DOI:

关键词: Unsupervised learning 、 Data mining 、 Artificial intelligence 、 Multiple kernel learning 、 Active learning (machine learning) 、 Statistical learning theory 、 Semi-supervised learning 、 Online machine learning 、 Machine learning 、 Instance-based learning 、 Computer science 、 Computational learning theory

摘要: This unique text/reference describes in detail the latest advances unsupervised process monitoring and fault diagnosis with machine learning methods. Abundant case studies throughout text demonstrate efficacy of each method real-world settings. The broad coverage examines such cutting-edge topics as use information theory to enhance tree-based methods, extension kernel methods multiple for feature extraction from data, incremental training multilayer perceptrons construct deep architectures enhanced data projections. Topics features: discusses frameworks based on artificial neural networks, statistical kernel-based methods; application steady state dynamic operations, a focus learning; spectral diagnosis.